The Empty Spreadsheet: Analyzing Vietnamese Volleyball When the Data Returns Zero

HOOK

On August 13, I reopened the 63-cell spreadsheet I had built for a domestic volleyball tournament. Nine dimensions, seven metrics each. Not a single cell contained a number.

The spreadsheet was not broken. It was empty because there was nothing to pour into it. Perfect first-pass rate: not published. Contacts per set: no log. Serve landing coordinates: nobody records them. The match sheet carried exactly three things — the score of each set, the winning team, and the list of scorers in each rally.

I sat in front of that screen for a while. Ten years earlier, as a second-year sociology student in Nha Trang, I was thrilled to scrape 380 Premier League matches from the 2026/2026 season in three weeks and compute PPDA for every team. Liverpool's 8.2 made the whole class laugh, until Liverpool reached Kyiv. Now I have a league forty kilometres from where I sit, and I cannot find one clean number to test my first assumption.

Data never lies, but it knows how to hide. In Vietnamese volleyball, it does not hide. It is simply absent.

CONTEXT

Vietnamese volleyball is at its highest peak of public interest in two decades. In 2026, the women's national team won SEA Games 31 gold on home soil in Hanoi. In 2026, they won the AVC Challenge Cup for the first time in Indonesia. Tran Thi Thanh Thuy, the captain, became the first Vietnamese female player to compete professionally abroad, moving to Japan and later Turkey. Nguyen Thi Bich Tuyen, standing over 1m85, became the name every opponent's analysis sheet circles in red.

Every big match produces thousands of comments, hundreds of edited clips, and very few numbers.

The paradox is this: interest grows exponentially while publicly available data stands almost still. I call this state the empty spreadsheet.

To be fair, the emptiness is structural, not accidental. The national women's league usually has about eight teams, plays a double round-robin split into two stages, and most matches are held in concentrated clusters in a single province over seven to ten days. That makes roughly fifty to sixty matches a season. Sixty matches a year is a small sample, but small samples can still be analysed, provided each match is deep enough. The problem is not the number of matches. It is the depth of each one.

Here is what actually exists. First: international community statistics sites, where data is user-entered, unverified, and limited to scores and a few aggregate metrics. Second: continental and world federation databases, open only at international events and never extended down to domestic league play. Third: the national league organiser's standings, publishing points, matches played, and occasionally per-match attack efficiency without set splits or rally counts. Fourth: scattered fan groups recording by hand, each with its own definition.

Four sources, four standards, none reconciling with the others. Vietnamese football has free play-by-play from global platforms, so PPDA or xG is a technical exercise. Vietnamese volleyball has no play-by-play, so everything downstream is decorative guesswork.

Based on my experience following these matches, both on screen and in person at the arena, the gap between the two sports is not about popularity. It is about recording habits. European football began logging every event more than twenty years ago and slowly standardised the definitions. Vietnamese volleyball has never entered that phase.

I am not writing this to complain. I am writing it because an empty spreadsheet is itself a finding. The absence of data measures a league's professionalisation more precisely than any press release.

CORE

The data architecture and where it breaks

Any volleyball analytics system rests on four layers. Layer one: the score, who scored, when. Layer two: per-rally behaviour — who touched the ball, how, and where it went. Layer three: position — where each player stood when the ball was served and when it was hit down. Layer four: context — the score, the set, substitutions, pressure level.

Vietnamese volleyball has layer one, lacks layer two, is nearly blank at layer three, and keeps layer four only in spectators' memory. Every advanced metric — opponent-adjusted attack efficiency, perfect first-pass rate over total receptions, blocking efficiency by system — requires layer two. No layer two, no metrics. Only anecdotes.

People still produce metrics. Many do it very well. But their input is still just the score sheet.

The small-sample trap: the number nine

A concrete example. An outside hitter plays a full four-set match. Her attack attempts land somewhere between eight and eleven. Call it nine. If she scores five, her efficiency is 55.6 percent. If she scores six, it is 66.7 percent. A single rally is worth 11.1 percentage points.

The 95 percent confidence interval for a 5-of-9 ratio runs from roughly 21 percent to 86 percent. With nine attempts, every efficiency number is statistically meaningless. Not noisy — meaningless. Narrowing that interval to a usable width requires about two hundred attempts per player. At nine attempts a match, that means more than twenty matches, close to two seasons, to produce one usable number for one player.

This is the technical reason every argument about who is better in Vietnamese volleyball ends in feeling. It is not that fans are lazy. The board is missing pieces.

Opponents are never adjusted for

Attack efficiency depends on the opponent's block height, the quality of the three-person block, and whether the opposing libero reads the setter well. The domestic women's league has wide gaps in class. A player facing the bottom half for five straight matches will post prettier efficiency than an equally good player facing the top half five times.

Without opponent data, those two numbers get placed side by side as if they shared a unit. The end-of-season individual standings therefore reflect the schedule more than they reflect ability. I am not saying this to dismiss the names at the top. I am saying their order has never been tested, and it may be right — by coincidence.

Blocking and system error

A block is credited to the last player to touch the ball. That means individual blocking metrics are really system metrics: who ran the block, who timed the setter, who read the ball. A successful block can be the work of the player who never touched it, the one whose positioning forced the attacker to hit cross-court into the blocker's hands. In the other direction, a failed block usually begins with a first pass that delivered the ball to an unwanted spot.

Layer two — who stood where, who moved how — is the only thing that clarifies this. Vietnamese volleyball has no layer two. So when a team loses three sets to nil, we describe it in feelings. When a middle blocker is praised for blocking well, we have no way to confirm it. This is the widest gap in the entire trade, wider even than the attack efficiency problem.

The season is long, the data is cold, and patience is the only measure.

Video: the real bottleneck

There is no data because nobody records it. Nobody records it because there is no tape. This is the point I want to stress most, because it is the most misunderstood.

Domestic volleyball is televised. But television is not an archive. The camera sits high and fixed, cutting constantly to follow the ball, with no wide angle showing six players at once, no downloadable file, no timestamps tied to individual rallies. A professional data recorder needs something very specific: a full-match video, one angle, uncut, scrubbable, timestamped.

Without that, recording has to happen live, from the stands, by eye. The error then exceeds the signal, and every conclusion comes with a disclaimer as long as the article itself.

What I actually have: four hundred hand-logged rallies

I tried once. I picked six sets of a second-stage women's match. I sat with a pre-drawn sheet and one hand on the keyboard. Roughly four hundred ball contacts in total, logged in five categories: serve, reception, set, attack, block. It took five hours. My self-declared error margin is about five percent, because in some rallies I was unsure who touched the ball last.

Four hundred rallies for one match. This is the only first-tier dataset I have, and I had to build it myself. In football, the equivalent data is free, verified, and definitionally standardised. I spent three weeks on three hundred and eighty football matches, and five hours on one volleyball match. The ratio of effort is one to fifteen thousand.

That ratio explains almost the entire quality gap between the two sports in Vietnam.

Transfers and playing abroad: where price has nothing to hold onto

When a foreign club considers a Vietnamese player, it has no dataset to reference. It has video, a few national team matches, and a handful of social media clips. That is the smallest sample of all small samples, and it is further biased toward keeping only the pretty rallies.

Transfers are not addition; they are the algorithm of greed. When the algorithm has no inputs, it runs on feeling. And feeling, at best, is a model with undefined coefficients.

Worth noting: we cannot answer the reverse question either. Whether Tran Thi Thanh Thuy suits the pace of the Turkish league is something nobody here can answer with numbers. We can only say she went, she played, she improved. All three statements are true, and none can be verified.

CONTRARIAN

Three things I find counter-intuitive.

First, more data does not automatically improve analysis. Expected goals has existed in football for over fifteen years, and every year I still see it used to prove the opposite of its original meaning. Data creates confidence, and confidence spreads faster than understanding. If the league organiser published full layer-two data tomorrow morning, the volume of wrong takes would rise, not fall, because many people would be holding a tool they were never trained to read.

Second, emptiness is itself an indicator. A league without per-rally logs is telling you it does not yet treat spectators as information partners, has no analytics department, and has no professional standard. These are organisational facts, measurable, and more useful than any score sheet. I recorded it in cell one: no data available.

Third, importing models without adjusting context is lying methodologically. Apply European attack-efficiency thresholds to a 1m75 outside hitter in the domestic league and the model will call her substandard, when the real issue is that the threshold belongs to a different environment, with different block heights and a different schedule density.

On the night Germany collapsed, I learned to test my own assumptions. On June 27, 2026, I tracked the kilometres run by German midfielders, the data said they would lose, and they lost to South Korea. I gave myself the pleasure of being right for a whole year, until I realised what I had actually proven: the data was right, but my context was wrong. A midfielder running little can signal decline, or it can signal a team that had less of the ball, or a player instructed to hold position. I had tested three of four assumptions and believed I had tested all four.

When the stadium is empty, the numbers begin to speak. In 2026, with no crowds, I had something close to a laboratory for measuring home advantage. The first results looked beautiful: home win rates fell sharply. But before publishing, I had to ask myself an uncomfortable question: how many other things changed at the same time. The schedule compressed, substitution allowances increased, and ranking criteria were adjusted. My laboratory had at least ten variables besides the crowd, not one.

Fans are not variables; they are weights. When the crowd is removed from the equation, what remains is not pure volleyball. It is volleyball minus one weight. And in Vietnam, we cannot even measure what we lost, because no national league match has ever been played in an empty arena with enough recording to compare.

There is something positive inside this deprivation. Vietnamese fans, lacking numbers to trade, are forced to argue with stories. Stories preserve emotion, collective memory, and those afternoons spent recalling a rally from ten years ago. A complete dataset would flatten that. But memory does not help when the decision is who makes the national team, who is overworked, and who deserves a move abroad. Memory answers questions about the past. Vietnamese volleyball needs answers about next season.

TAKEAWAY

I will track three signals over the next twelve months. One: whether the national league organiser publishes per-rally logs in at least one division. Two: whether any platform does touch-by-touch logging, even for a handful of featured matches. Three: whether I myself accumulate two hundred hours of manual logging to build a usable sample for three players.

Until then, I leave cell one blank and write beside it: no data available. That is the most honest sentence a spreadsheet can produce.

Before you burn the tactics board, check your data source. If the source is a blank page, the plan is not wrong — it is simply untested. And an untested plan, at the SEA Games, at the AVC, or in a provincial arena on a Saturday afternoon, carries the same unmeasurable probability of failure. I do not believe in instinct; I believe in the moment instinct gets digitised. In Vietnamese volleyball, that moment has not arrived.